AncestryDNA results

Question

My uncle Erskine have traced our family tree years ago and gave us all copies of it. Nikita and trader Hughes obviously was a part of it. I took my DNA test through ancestry in December and it seems like everything that he and family research matches with my DNA. If you would like to have a look at it and see my ancestry DNA in the names that are attached to it you’re more than welcome to see. I’m trying to understand how DNA works. Because I don’t show any Indian indigenous DNA in my test for me. But my family tree has a lot of indigenous people in it as well as African-Americans. And the names on the list on the same list that match up to my DNA ancestors through my ancestry DNA connections

Answer 1

It''s not possible for autosomal DNA to connect anyone to someone who lived 400 years ago.  At 10 generations back you have over 1000 ancestors, so the odds of your inheriting any noticible DNA from one of them is infinitesimal. Your DNA may connect you to other people who are related to you and with whom you share a common ancestor in the last five or six generations.   There is no documentation which supports the existence of anyone named "Nicketti" in the Jamestown period when she supposedly lived.

Comments

Your information regarding the limits of autosomal DNA for verifying relationships is incorrect. It can certainly verify relationships more distant than 6 generations. My most distant confirmed autosomal DNA matches share common ancestors who were born in the mid-1600s.

---

Autosomal DNA might connect you to living people who are also descended from someone born 400 years ago - the mythical Nicketti is supposed to have been born about 1625 - but I don’t think there’s any way to connect your DNA today with a specific person that long ago unless you find him or her, get DNA from the remains, and match to a now-living person as they did with  Richard III.

Answer 2

I agree fully with Kathie's main point. But just to disagree with a detail or two:

1. The big question is: how many generations back do you have Nicketti Powhatan in your tree? Her birth year is only a proxy for generational difference, but, for instance, I have a 4th-great grandfather who was born in 1746. There are a couple of men who are reportedly still living as of 2019 who are grandchildren of a man born in 1790 (that man being US President John Tyler). So you can get pretty far back without going through too many generations if you are lucky.

2. Let's say we are talking about 9 generations back -- close to the 10 Kathie mentioned -- so we're talking about 7th great-grandparents. The amount of DNA shared with a 7th great-grandparent is the same as that expected between 4th cousins. That is, you would share typically about 13cM, so a detectable amount ... if you could test your 7th great grandparent, but unless you have a paleogenetics lab of your own, this is all hypothetical. You'll share more than this with some 7th-great-grandparents and less than others, but it's estimated also that you'd have a little over 50% chance of sharing detectable DNA with a typical 7th-great grandparent. So saying "infinitesimal", even at one generation further back, is an exaggeration.

Given another 9-generation descendant of that 7th great-grandparent, there is then about a 25% chance that both you and that cousin share detectable DNA with that ancestor. The problem is that it is then exceedingly unlikely that you share the same piece of DNA. It may be, for instance, that you share the first part of chromosome 7 with your 7th-great grandparent while this other cousin just shares a little chunk about 2/3 the way along chromosome 18. So you didn't inherit the same piece of DNA, and usually when that happens, you have no other shared DNA either -- you won't be matches.

Fortunately, if you share DNA with an 7th great-grandparent, typically you will have thousands of living cousins who descend from that same person. If you share a 13cM segment with the 7th great-grandparent, then because you can fit only a few hundred 13cM segments into your wholegenome, the chances aren't terrible that one of those cousins will also have inherited a detectable segment that significantly overlaps your 13cM segment.

But will that cousin have tested? Probably not. If they have, will they have published their tree? Probably not. If they have published their tree, will it extend back and include that 8th great-grandparent? Probably not.

But you can repeat this same game with your 1024 eigth great grandparents, and there is probably a non-infinitesimal chance that you can confirm one 8th cousin with a segment inherited by descent. But finding that person is like finding a needle in a haystack. I've found lots of individual matches to 8th cousins and beyond, but none has yet triangulated and I assume that they are all false matches, or IBP, or at least may have come down a line other than the one I think -- maybe with a more recent common ancestor.

Back to Jennifer: you want to use autosomal DNA to confirm descent from a *specific* ancestor rather than the much weaker goal of "some one of my many ancestors back 9 or so generations." That is exceedingly unlikely. Maybe if a thousand people played this same game, one could get lucky. But chances are, you are not that person (and neither am I).

For me -- I have confirmed lots of 5th cousins through individual matches, and most of those with good trees indicate there is no other reasonable place for the shared DNA to have come from than the ancestor we identified. Past this, it is still possible to confirm 6th cousins, maybe 7th, and maybe even 8th, but difficulty goes up exponentially with each generation back, and your probability of success goes down. But if you are one of the lucky people that can confirm an eighth cousin with a nice solid triangulation, know that you won't get to pick the ancestor you confirm in this manner. It will just be luck of the draw.

Comments

My uncle uncle did a very detailed family tree. My AncestryDNA test relates me Davis Floyd’s Ralfs so many more. My other question I was wondering how much DNA do we share from our 2nd or 3 and 4th grandparents

---

You share about 6.3% with a 2nd great-grandparent, 3.1% with a 3rd, and 1.6% with a 4th. That'sabout 425cM, 53cM, and 13cM.

But there is huge variation in these numbers. You could share anywhere between 0 and 150cM with a 3rd great-grandparent, maybe more. 53cM just just an average expected amount.

When you look at ethnicity estimates, below 6% or so and people start to call it "trace". It becomes much less trustworthy. So unless you have around 2nd-great-grandparent who was full Native American, it probably won't show up in your ethnicity estimate, and if it does it will be hard to tell if it is real or just a false positive.

---

This also is not true? Continental-level populations are very distinct from each other due to many thousands of years of geographic separation. Native American DNA is distinct enough from European DNA to be identified with a high degree of certainty. And six percent is not a "trace" (a fractional percentage may be a trace and due to a false positive; over 1% identified as being from a distinct continental population is extremely unlikely to be due to a false positive).

---

"Also?" Is there something else I said that you think is untrue?

You say trace is <1%. I said "people start to call it 'trace'". i.e., at that point some of the people are saying trace, and others are not. There is no hard line to define trace -- 1% is no more a hard line than 6%. It's like saying a p-value of 4.9 is "significant" but 5.1 is not. Why is 5 special? (It isn't)

But otherwise, I never said anything about not being able to distinguish continent-level populations. Indeed, my answer made it clear, I thought, that is possible if you use a reasonably close ancestor, like a 2nd-great-grandparent.

---

People referring to <6% as "trace" are usually "white" Americans who are upset at not being 100% European trying to ignore their non-European ancestry. You can also inherit detectable DNA from ancestors much more distant than 2nd great-grandparents. There are other Wikitree users I share DNA with whose closest relationship is 7th or 8th cousin. And in the case of Y chromosome DNA or mitochondrial DNA, it is very possible to identify an ancestor from a distinct continental population despite having zero autosomal DNA from that population.

---

There's a huge difference between inheriting detectable amounts of shared DNA and inheriting enough DNA to be confidently checked against a certain geographic reference population. My post made it clear that you will inherit detectable DNA from most of your 7th-great-grandparents. 

For the rest:

• the OP was not talking about non-recombinant DNA
• the first statement -- if you're going to speculate, it is more common in my experience to see people getting upset at the opposite situation, for the opposite reason.

Answer 3

Jennifer, thank you for being brave enough to come out into the g2g and ask your question.

Instead of going into a bunch of scientific language, it is easiest to say that:

You receive half of your autosomal DNA from your mother and the other half from your father.

Your mom and dad each get 1/2 of their autosomal DNA from their mother and the other half from their father.

So you receive 1/4 of your autosomal DNA from your grandmother and another 1/4 from your grandfather.

The number of specific DNA markers you have is reduced by 50% every generation back you go.

So when you get to the 8th generation you receive only 1/256th, which is much less than 1% of autosomal DNA.

The 10th generation ancestor shares less than 1/1000 of autosomal DNA with us.  That makes it very difficult to read with any accuracy.

I hope this helps you understand. Have a great weekend.